Article
Engineering, Civil
Jui-Liang Lin, Ming-Chieh Chuang
Summary: Simplified nonlinear modeling is useful for estimating the seismic response of buildings and has the advantages of providing structural perception and computational efficiency. The present study enhances the generalized building model (GBM) by developing the inelastic properties of the sticks for nonlinear response history analysis of buildings.
ENGINEERING STRUCTURES
(2023)
Article
Acoustics
Wei Xu, Min Xu, Xiaomin An, Weigang Yao
Summary: The study presents a nonlinear reduced-order model for calculating the nonlinear response of isotropic and composite plates, showing nearly an order of magnitude speedup compared to direct FE simulation and predicting shorter sonic fatigue life than White Gaussian Noise (WGN).
JOURNAL OF SOUND AND VIBRATION
(2021)
Review
Chemistry, Multidisciplinary
Shilei Dai, Xu Liu, Youdi Liu, Yutong Xu, Junyao Zhang, Yue Wu, Ping Cheng, Lize Xiong, Jia Huang
Summary: Living organisms possess a mysterious and powerful sensory computing system based on ion activity. The development of iontronic devices in recent years has provided a promising platform for simulating the sensing and computing functions of living organisms. These devices can generate, store, and transmit signals by adjusting ion concentration and distribution, bridging biology and electronics, and offering advantages in sensing and recognition. This review provides an overview of neuromorphic sensory computing by iontronic devices, highlighting concepts and breakthroughs, and discusses challenges and future directions.
ADVANCED MATERIALS
(2023)
Article
Chemistry, Multidisciplinary
Geunyoung Kim, Seoil Son, Hanchan Song, Jae Bum Jeon, Jiyun Lee, Woon Hyung Cheong, Shinhyun Choi, Kyung Min Kim
Summary: A Pt/Ta2O5/Nb2O5-x/Al2O3-y/Ti CTM stack with high retention and array-level uniformity is proposed as a reliable selector-less MCA. It exhibits self-rectifying and nonlinear current-voltage characteristics, making it suitable for non-volatile analog applications.
Review
Chemistry, Physical
Shi-Jun Liang, Yixiang Li, Bin Cheng, Feng Miao
Summary: In recent decades, advances in material science have allowed for precise control of material structures at the atomic scale and the realization of various low-dimensional heterostructures. These heterostructures have been widely used in electronic, spintronic, and optoelectronic devices with great success. Manipulating the dimensionality and physical properties of low-dimensional materials enables the design of a wide range of low-dimensional artificial heterostructures. The unique physical properties exhibited by these emerging low-dimensional heterostructures provide unprecedented opportunities for neuromorphic computing applications, laying the groundwork for future intelligent systems.
Article
Engineering, Multidisciplinary
Saurabh Balkrishna Tandale, Marcus Stoffel
Summary: The present study aims to introduce an AI algorithm suitable for neuromorphic computing to solve Boundary Value Problems in Engineering Mechanics. By using Spiking Neural Networks (SNNs), the study proposes a surrogate model for mechanical tasks that is more energy-efficient than traditional neural networks. The researchers also propose a hybrid model that combines spiking recurrent cells, the spiking variant of the Legendre Memory Unit (LMU), and classical dense transformations to compute the nonlinear response of shock wave-loaded plate elements.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
Article
Astronomy & Astrophysics
J. J. A. Baselmans, F. Facchin, A. Pascual Laguna, J. Bueno, D. J. Thoen, V Murugesan, N. Llombart, P. J. de Visser
Summary: Future space-borne observatories with passive cooling can potentially achieve high sensitivity for far-infrared observations using MKIDs. Optimization of MKID geometry and reduction of excess noise effects can result in sufficient sensitivity for spectroscopic observations.
ASTRONOMY & ASTROPHYSICS
(2022)
Article
Engineering, Civil
Soheil Sadeghi Eshkevari, Martin Takac, Shamim N. Pakzad, Majid Jahani
Summary: A physics-based recurrent neural network model is proposed in this study to accurately estimate dynamic responses of linear and nonlinear multi-degree-of-freedom systems. Compared with other models, this model has higher accuracy and requires fewer trainable variables. Numerical case studies demonstrate the network's ability to learn different nonlinear behaviors of dynamic systems with high accuracy.
ENGINEERING STRUCTURES
(2021)
Article
Optics
Jacek Gosciniak, Zibo Hu, Martin Thomaschewski, Volker Sorger, Jacob B. B. Khurgin
Summary: This article proposes a novel bistable resonator-free all-optical waveguide device based on indium tin oxide as a nonlinear epsilon-near-zero material. The device offers a cost-effective and high-performance photonic platform for optical logic and information processing in the next generation optical networks and photonic neural systems. It is compatible with silicon photonics, enabling sub-picosecond operation speeds with moderate switching power. The device can act as an optical analogue of memristor or thyristor and can become an enabling element of photonic neural networks not requiring OEO conversions.
LASER & PHOTONICS REVIEWS
(2023)
Article
Chemistry, Multidisciplinary
Xiang Wan, Tohru Tsuruoka, Kazuya Terabe
Summary: Center-surround antagonism, a key mechanism in the retina, is successfully emulated in a neuromorphic system using multiple ionic devices with lithium cobalt oxide channels arranged on a common electrolyte. By exciting single devices and implementing device-to-device inhibition, the system is able to encode edge contrast and achieve edge detection for real images. The simplicity and effectiveness of this system is attributed to the intrinsic properties of the materials employed.
Article
Chemistry, Multidisciplinary
Qiyu Yang, Zheng-Dong Luo, Dawei Zhang, Mingwen Zhang, Xuetao Gan, Jan Seidel, Yan Liu, Yue Hao, Genquan Han
Summary: This paper demonstrates a van der Waals (vdW) heterostructure-based optoelectronic transistor that can integrate photoreceptor, memory, and computation functions. It exploits diverse photoelectric control to achieve versatile photoresponse characteristics and can emulate short- and long-term synaptic plasticity and psychological human memory. This promising hardware system for visual information in-sensor computing offers potential solutions for complex computation tasks.
ADVANCED FUNCTIONAL MATERIALS
(2022)
Article
Computer Science, Artificial Intelligence
Akhil Bonagiri, Dipayan Biswas, Srinivasa Chakravarthy
Summary: The development of nonlinear dynamical electronic devices and circuits has made it possible to implement energy-efficient hardware realizations of neurobiological systems. Central pattern generator (CPG) is a neural system that controls rhythmic motor behaviors in animals. A compact and energy-efficient hardware platform for implementing neuromorphic CPGs would greatly benefit bio-inspired robotics. In this study, we demonstrate that four capacitively coupled VO2 memristor-based oscillators can produce spatiotemporal patterns corresponding to quadruped gaits and show similarities to conductance-based neuron models.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Review
Chemistry, Multidisciplinary
Kaixuan Sun, Jingsheng Chen, Xiaobing Yan
Summary: This article reviews the applications and development of memristors in the field of neural networks, pointing out that memristor architecture is expected to become an alternative to the von Neumann architecture and address the challenges of the neural network and big data era.
ADVANCED FUNCTIONAL MATERIALS
(2021)
Article
Mathematical & Computational Biology
Aditi Anand, Sanchari Sen, Kaushik Roy
Summary: Quantifying the similarity between artificial neural networks (ANNs) and biological counterparts is crucial for building more brain-like artificial intelligence systems. Recent research shows that a non-linear mapping function can lead to higher neural predictivity, but improvements in classification performance of image recognition ANNs do not necessarily translate to better neural predictivity.
FRONTIERS IN COMPUTATIONAL NEUROSCIENCE
(2021)
Article
Astronomy & Astrophysics
H. Denes, K. M. Hess, E. A. K. Adams, A. Kutkin, R. Morganti, J. M. van der Hulst, T. A. Oosterloo, V. A. Moss, B. Adebahr, W. J. G. de Blok, M. Ivashina, A. H. W. M. Coolen, S. Damstra, B. Hut, G. M. Loose, D. M. Lucero, Y. Maan, A. Mika, M. J. Norden, L. C. Oostrum, D. J. Pisano, R. Smits, W. A. van Cappellen, R. van den Brink, D. van der Schuur, G. N. J. van Diepen, J. van Leeuwen, D. Vohl, S. J. Wijnholds, J. Ziemke
Summary: This study aims to characterize the response of the Apertif compound beams in terms of frequency and polarization. The measured beam maps can be used for various data processing steps, such as image deconvolution and primary beam correction. The results show good agreement between the beam shapes derived using different methods.
ASTRONOMY & ASTROPHYSICS
(2022)
Article
Computer Science, Hardware & Architecture
Xiangwei Li, Douglas L. Maskell, Carol Jingyi Li, Philip H. W. Leong, David Boland
Summary: This article presents an efficient FPGA implementation of the FFT accumulation method (FAM) for estimating the spectral correlation density (SCD) function and its alpha profile. The implementation achieves a high PE utilization of 88.2% using a linear systolic array with bidirectional datapath consisting of DSP-based processing elements. It consumes less than 36% of the logic fabric on a Zynq UltraScale+ XCZU28DR-2FFVG1517E RFSoC device and has significantly better energy efficiency compared to GPU and hybrid FPGA-GPU implementations.
ACM TRANSACTIONS ON RECONFIGURABLE TECHNOLOGY AND SYSTEMS
(2023)
Article
Nanoscience & Nanotechnology
Iat Wai Leong, Makusu Tsutsui, Kazumichi Yokota, Sanae Murayama, Masateru Taniguchi
Summary: By using partial dielectric coatings, the nonlinear ionic current through a pore can be altered, allowing the pore to behave like a resistor, diode, and bipolar junction transistor. The reasons for asymmetric ion transport in the pore and the relationship between specifically charged surfaces and electroosmotic flow are revealed through numerical simulations. These findings provide a direct approach to modify the electroosmotic-flow-driven ionic current rectification in channel-based devices via dielectric coatings.
ACS APPLIED MATERIALS & INTERFACES
(2023)
Review
Chemistry, Multidisciplinary
Arman Ahnood, Andre Chambers, Amy Gelmi, Ken-Tye Yong, Omid Kavehei
Summary: The advent of electronic technology for neural interfacing in the past 50 years has transformed medicine and biology. Devices like deep brain stimulators and cochlear implants have revolutionized the treatment of previously untreatable conditions. Semiconducting electrodes have advanced neural interfacing technologies, enabling high precision sensing and stimulation in electrical, biochemical, and optical domains. This emerging class of electrodes offers new opportunities for research and treatment.
CHEMICAL SOCIETY REVIEWS
(2023)
Article
Mathematics, Interdisciplinary Applications
Weijie Xu, S. C. Chan, W. Y. Leong
Summary: Artificial intelligence plays a decisive role in the healthy and sustainable development of society, particularly in fire supervision. The application of intelligent fire control enables the fire department to better supervise fire control and enhance the pertinence and effectiveness of fire control supervision.
DISCRETE DYNAMICS IN NATURE AND SOCIETY
(2023)
Article
Computer Science, Hardware & Architecture
Yansong Gao, Jianrong Yao, Lihui Pang, Wei Yang, Anmin Fu, Said F. Al-Sarawi, Derek Abbott
Summary: To improve the modeling resilience of silicon strong PUFs, various composited APUF variants have been devised. However, there is a challenge in evaluating their modeling resilience using multiple information sources. This paper proposes a deep learning attack called MLMSA to thoroughly evaluate the resilience of APUF variants, which can successfully break large-scaled APUF variants and is a useful technique for evaluating other PUF designs.
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS
(2023)
Article
Engineering, Environmental
Bright Brailson Mansingh, Joseph Selvi Binoj, Ze Quan Tan, Wai Leong Eugene Wong, Taweechai Amornsakchai, Shukur Abu Hassan, Kheng Lim Goh
Summary: This study aims to develop a sustainable and eco-friendly composite material for food product packaging using existing production technology. Chitin and chitosan were blended with polylactic acid (PLA) pellets, and after extrusion, the composite parts were subjected to various tests and analyses. The addition of chitin and chitosan resulted in reduced tensile and flexural properties but improved ductility. Microscopic examination revealed the presence of voids and decomposed chitin and chitosan. Fourier transform infrared spectra analysis provided insights into the chemical bonding, while X-ray diffraction analysis showed the crystal size and crystallinity index. Lastly, thermogravimetric analysis confirmed the thermal stability of the composite materials. These findings support the use of Chitosan/PLA and Chitin/PLA composites in food product packaging.
JOURNAL OF POLYMERS AND THE ENVIRONMENT
(2023)
Review
Computer Science, Artificial Intelligence
Andrew C. Cullen, Benjamin I. P. Rubinstein, Sithamparanathan Kandeepan, Barry Flower, Philip H. W. Leong
Summary: The growth in devices accessing the wireless spectrum has increased due to the Internet of Things and 5G. As a result, there has been a corresponding increase in spectrum conflict and congestion, impacting innovation in both public and private sectors. Dynamic Spectrum Allocation, which involves devices making intelligent decisions about access to spectrum, could potentially resolve these issues and improve efficiency. However, this would require the development of complex inference frameworks and tools to simulate, measure, and predict Spectral Occupancy.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Computer Science, Hardware & Architecture
Huming Qiu, Hua Ma, Zhi Zhang, Yansong Gao, Yifeng Zheng, Anmin Fu, Pan Zhou, Derek Abbott, Said F. Al-Sarawi
Summary: This article proposes a reconfigurable 1-bit quantized deep binary neural network (RBNN) that achieves efficient memory utilization and model IP protection on IoT devices.
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS
(2023)
Article
Engineering, Biomedical
Zhangyu Xu, Nhan Duy Truong, Armin Nikpour, Omid Kavehei
Summary: This study introduces a proof-of-concept optical telemetry module that utilizes a single LED for high-speed data transmission with low power consumption and small size. The experiments demonstrated data rates of 108 Mbit/s and 54 Mbit/s for tissue thicknesses of 3 mm and 8 mm, respectively. The module is powered by near-field coupling and achieves bidirectional communication through low-speed downlink from near-field communication, aiming to provide reliable transmission in a miniaturized implant for high-speed wireless communication.
JOURNAL OF NEURAL ENGINEERING
(2023)
Article
Computer Science, Information Systems
Lachlan Burne, Chiranjibi Sitaula, Archana Priyadarshi, Mark Tracy, Omid Kavehei, Murray Hinder, Anusha Withana, Alistair McEwan, Faezeh Marzbanrad
Summary: This article proposes a novel technique for automated peristalsis sound detection from neonatal abdominal sound recordings and compares it to various other machine learning approaches. It adopts an ensemble approach that utilizes handcrafted as well as one and two dimensional deep features obtained from Mel Frequency Cepstral Coefficients (MFCCs). The results show that our method provides an accuracy of 95.1% and an Area Under Curve (AUC) of 85.6%, outperforming both the baselines and the recent works significantly. These encouraging results demonstrate that our proposed Ensemble-based Deep Learning model is helpful for neonatologists to facilitate tele-health applications.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Review
Multidisciplinary Sciences
Christina Maher, Yikai Yang, Nhan Duy Truong, Chenyu Wang, Armin Nikpour, Omid Kavehei
Summary: Epilepsy is a common condition characterized by recurrent, unpredictable seizures. The current method of monitoring with EEG is time-consuming, uncomfortable, and sometimes ineffective for patients. Limited options for data collection and training machine-learning models exist due to restricted hospital resources and hardware/software specifications. This mini-review looks at the current patient journey and suggests opportunities for improving data reliability through multi-modal data fusion. Further research is needed to advance brain monitoring solutions towards portable, reliable devices that offer patient comfort, perform ultra-long-term monitoring, and expedite the diagnosis process.
ROYAL SOCIETY OPEN SCIENCE
(2023)
Article
Food Science & Technology
M. R. Abd Rahman, Z. Hassan, M. S. Hassan, R. Hashim, L. S. Wong, W. Y. Leong, S. H. Syd Jaafar, S. Salvamani
Summary: This study aimed to investigate the effect of supplementing dairy goats with date pit powder (DPP) on the unsaturated fatty acid (FA) content in milk. The results showed that higher doses and longer supplementation time of Mariami DPP can enhance the quantity and quality of FA in milk.
INTERNATIONAL JOURNAL OF FOOD SCIENCE
(2023)
Article
Biophysics
Veronika Timosina, Tim Cole, Hongda Lu, Jian Shu, Xiangbo Zhou, Chengchen Zhang, Jinhong Guo, Omid Kavehei, Shi-Yang Tang
Summary: Biopotential signals, like ECG, EMG, and EEG, are important for diagnosing various disorders. Traditional dry electrodes have issues with movement and impedance imbalance, while conductive hydrogels dry over time. To overcome these problems, a non-eutectic Ga-In alloy as a shear-thinning non-Newtonian fluid is used to create superior electrodes.
BIOSENSORS & BIOELECTRONICS
(2023)
Article
Computer Science, Information Systems
Hua Ma, Qun Li, Yifeng Zheng, Zhi Zhang, Xiaoning Liu, Yansong Gao, Said F. Al-Sarawi, Derek Abbott
Summary: The use of cryptographic privacy-preserving techniques in Federated Learning (FL) creates a security dilemma as tampered local model parameters are encrypted and not auditable. This research demonstrates the ease of performing model corruption attacks in privacy-preserving FL. A protocol called MUD-PQFed is proposed to detect and penalize malicious attacks effectively, preserving the utility of the global model by removing contributions from detected malicious clients.
COMPUTERS & SECURITY
(2023)
Article
Engineering, Electrical & Electronic
Yikai Yang, Jason K. Eshraghian, Nhan Duy Truong, Armin Nikpour, Omid Kavehei
Summary: This study proposes a novel approach for seizure detection using a deep spiking neural network (SNN) based on neuromorphic computing. The proposed SNN model achieves high accuracy on publicly accessible datasets, including CHB-MIT, FB, and EPILEPSIAE datasets, with average cross-validation area under the curve scores of 92.7%, 89.0%, and 81.1% respectively. The computational overhead and energy consumption of the proposed model are significantly reduced compared to alternative state-of-the-art models, indicating the potential for building an accurate and low-power neuromorphic system. This study is the first feasibility study using a deep SNN for seizure detection on reliable public datasets.
NEUROMORPHIC COMPUTING AND ENGINEERING
(2023)